Algorithmic opaqueness to financial disengagement: buffering role of community well-being in AI anxiety and perceived redlining across rural regions in the Global North

Purpose This study investigates how algorithmic opacity (ALOP) and inadequate AI governance influence perceptions of digital redlining (PRDL), financial anxiety (FNAX) and trust erosion, leading to withdrawal from traditional banking. The paper further explores how community well-being and empowerment may mitigate these adverse impacts among residents of remote regions in North America and Europe. Design/methodology/approach Study 1 followed a sequential mixed-methods design, in which 32 in-depth interviews were conducted across eight peripheral areas in the Global North: Nunavut, Yukon, Appalachia, Northern Maine, the Highlands and Islands of Scotland, Lapland, Galicia and the Eastern Carpathians. Study 2 surveyed 575 participants to test a Technology Threat Avoidance Theory (TTAT)–Stimulus-Organism-Response (S-O-R) model through confirmatory factor analysis and structural equation modelling, examining psychological pathways and moderating effects. Findings Results show that opaque algorithms and poor governance potentially amplify PRDL and FNAX, eroding institutional trust and promoting withdrawal behaviour. However, community empowerment can weaken these effects. In well-connected, digitally literate communities, residents view algorithmic shortcomings as a systemic issue rather than a personal failure, thereby maintaining trust and engagement through shared learning and local collaboration. Originality/value The study extends TTAT and S-O-R frameworks to AI-mediated financial services and offers a cross-regional perspective on AI's psychological impact in marginal contexts. It demonstrates that sustainable digital inclusion depends not only on transparent algorithms but also on human-centred governance and empowered communities, highlighting collective resilience as a practical countermeasure to ALOP and financial exclusion.